Weekly Carboplatin and Paclitaxel: A Retrospective Comparison with the Three‐Weekly Schedule in First‐Line Treatment of Ovarian Cancer

Abstract

Background

Conventional first‐line combination therapy for ovarian cancer comprises 6 cycles of adjuvant or neoadjuvant carboplatin (AUC5‐6) with paclitaxel (175 mg/m²) every 3 weeks (PC‐3W). Weekly scheduling of paclitaxel may maximize its antiangiogenic effect and reduce adverse effects. We compared the efficacy and safety of PC‐3W with a modified protocol of weekly paclitaxel 80 mg/m² and weekly carboplatin AUC2 administered on days 1, 8, and 15 in a 28‐day cycle (i.e., with 1 week off‐treatment [PC‐W]).

Materials and Methods

Medical records of consecutive patients treated between 2000 and 2018 were reviewed; 707 patients were analyzed for demographic and clinical characteristics, effectiveness and toxicity.

Results

PC‐3W was administered to 402 patients (median age, 60.5 years) and PC‐W to 305 patients (median age, 62.5 years). Most patients (91.4%) were diagnosed at stage III–IV. Notwithstanding a higher proportion of residual disease and older patients in the PC‐W group, median progression‐free survival was 21.4 months and 13.2 months for PC‐W and PC‐3W, respectively; median overall survival was 75.2 and 54.0 months for PC‐W and PC‐3W, respectively. Cox proportional hazards model indicated improved survival for patients treated with PC‐W (hazard ratio, 0.54). Similar results were observed for older patients diagnosed at ≥75 years. PC‐W demonstrated a better safety profile, with lower incidence of neuropathy, neutropenia, and alopecia.

Conclusion

PC‐W is as active and better tolerated than the standard PC‐3W regimen. PC‐W may serve as an alternative option for elderly or frail patients.

Implications for Practice

Weekly scheduling of paclitaxel 80 mg/m² and carboplatin AUC2, administered on days 1, 8, and 15 in a 28‐day cycle (PC‐W) for first‐line therapy for advanced ovarian cancer, is as active and better tolerated than the standard regimen of carboplatin and paclitaxel (175 mg/m²) every 3 weeks (PC‐3W). It is possible that the weekly holiday on day 21 in the PC‐W regimen may ensure better completion rates (which may result in treatment delays for toxicity in PC‐3W). The results of this retrospective analysis highlight the weekly regimen as a valid treatment option, especially for elderly patients and those with significant comorbidities.

Introduction

Epithelial ovarian cancer (EOC) is the leading cause of gynecological cancer mortality [1, 2]. The conventional, and most studied, first‐line combination therapy for EOC consists of carboplatin (AUC5‐6) with paclitaxel (175 mg/m²) repeated every 3 weeks for six cycles either as an adjuvant or neoadjuvant administration. In randomized trials comparing this regimen with others, the 3‐weekly regimen was superior or noninferior to other regimens in terms of response rate, progression‐free survival (PFS), and overall survival (OS) [3–6]. Its adverse events include alopecia, neurotoxicity, myelosuppression, and fatigue [6, 7].

Paclitaxel binds to tubulin and inhibits the disassembly of microtubules, thereby resulting in the inhibition of cell division [8, 9]. Weekly scheduling of paclitaxel has been suggested to increase its efficacy by reducing tumor regrowth between cycles and maximizing the drug's antiangiogenic effect [10, 11]. Weekly administration also reduces acute myalgia and arthralgia, presumably by lowering peak plasma concentrations [12]. Accordingly, several trials have explored paclitaxel dose schedules at shorter intervals [13–16]. The Japanese Gynecologic Oncology Group (JGOG) reported remarkable superior efficacy in their “dose‐dense protocol” consisting of weekly paclitaxel (80 mg/m²) with carboplatin (AUC6) administered every 3 weeks versus the standard 3‐weekly administration [15]. However, subsequently the Gynecologic Oncology Group (GOG) failed to show such benefit in the GOG262 study that revisited such a comparison but included a percentage of patients with optional bevacizumab randomized to both arms [13]. A third study (MITO 7) reported similar effectiveness but fewer side effects for a weekly protocol consisting of carboplatin AUC2 with paclitaxel 60 mg/m² compared with the standard 3‐weekly protocol [16]. Therefore, after nearly universal adoption of weekly paclitaxel regimens after the JGOG publications, the trend for induction regimens has moved back to ones that are administered every 3 weeks.

Since our phase II study of 64 patients from 2003 to 2007 evaluating the safety and efficacy of a modified double protocol [17], some of our physicians have continued to use this regimen, which we define as PC‐W. It consists of carboplatin (AUC2) and paclitaxel (80 mg/m²) administered on days 1, 8, and 15 in a 28‐day cycle (PC‐W). In the current study, we retrospectively compared the activity and safety of the PC‐W protocol to the standard 3‐weekly protocol (PC‐3W) in a large cohort of unselected patients with advanced EOC.

Materials and Methods

Setting and Patients

Medical records of all consecutive patients with advanced EOC, tubal carcinoma (TC), or primary peritoneal carcinomatosis (PPC) treated between 2000 and 2018 at the Onco‐Gynecological Unit of Tel Aviv Medical Center (629 cases, 89.0%) and at other medical centers (78 cases, 11.0%) were reviewed retrospectively.

All patients treated with a carboplatin‐paclitaxel regimen as a first‐line treatment (either PC‐W or PC‐3W) were included in the analysis. The first 64 patients on weekly schedule were those that were also included in the prospective, single‐arm, phase II study previously published [17].

The collected data included demographic and clinical characteristics, as well as survival status and toxicities.

Treatment Schedule

Patients underwent either primary surgical debulking or interval debulking after three to four neoadjuvant chemotherapy cycles. The 3‐weekly regimen consisted of the conventional doublet paclitaxel 175 mg/m² and carboplatin AUC6, on day 1 of a 21‐day cycle repeated for 6 cycles (PC‐3W). The weekly regimen consisted of paclitaxel 80 mg/m² and carboplatin AUC2 administered on days 1, 8, and 15 of a 28‐day cycle, repeated for six cycles (PC‐W). It is important to note that some patients with R2 and R1 residual disease received more than six cycles.

The choice to administer the PC‐W regimen was often based on the patient's performance status (i.e., older women or women with advanced diseases and especially severe ascites). In addition, the regimen was offered to women who preferred having severe grade 1 alopecia compared with grade 2 alopecia (i.e., complete hair loss). An additional requirement was the ability to come to the medical center every week to receive treatment. The physician explained to the patients that in phase II studies the PC‐W regimen showed similar effectiveness to the PC‐3W regimen, as well as lower toxicities (specifically alopecia and neuropathy), but that this was not confirmed in larger phase III studies.

Assessment of Outcomes and Safety

Serum cancer antigen (CA)‐125 levels were evaluated every three cycles during chemotherapy treatment. Complete blood count was performed before each treatment. A confirmed increase in serum CA‐125 levels to more than twice the upper limit of normal values (35 U/dL) was considered as disease progression for patients with no corresponding imaging evidence of disease.

OS was calculated from the initiation of treatment with carboplatin and paclitaxel to either death or to the last known follow‐up. PFS was calculated from the last treatment with carboplatin and paclitaxel to either progression or death or to the last known follow‐up. Platinum sensitivity was defined as recurrence of disease more than 6 months after the last platinum‐based treatment.

At the end of each treatment cycle, toxicity was evaluated according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 2. At each follow‐up visit (for the duration of patient follow‐up) we also asked the patient if she had peripheral sensory neuropathy.

Statistical Analysis

For each treatment group, continuous variables were summarized as median and range, and categorical variables were summarized as number and percentage. Survival functions were demonstrated using the Kaplan‐Meier method, and the effect of each treatment group was assessed using the Log‐Rank test. Univariate comparisons between groups were performed using either the independent Student's t test or the Mann‐Whitney U test for continuous variables and Fisher's exact test or chi‐squared test for categorical variables.

The Cox proportional hazards model was used to compare the risk of death and the risk of a composite endpoint of death or recurrence between the two treatment regimens, adjusting for age, stage of disease, histology (serous papillary, adenocarcinoma, other), BRCA 1 and BRCA 2 status (carriers vs. noncarriers), and debulking status. The model for OS was also adjusted for platinum sensitivity (i.e., platinum‐free interval > 6 months). Time to recurrence and/or death was defined as the time from diagnosis until the relevant endpoint. Results are presented as the hazard ratio (HR) and the 95% confidence interval (CI). The level of significance used for all of the above analyses was two‐tailed, and a p value of <.05 was considered significant. Statistical analyses were performed using R software, version 3.6 (R Development Core Team, Vienna, Austria).

Results

Between 2000 and 2018, 707 patients with EOC, TC, or PPC were treated with first‐line PC‐3W (402, 56.9%) or PC‐W (305, 43.1%). Baseline characteristics are shown in Table 1. The median age of diagnosis of the patients treated with PC‐W was higher than the age of those treated by PC‐3W (62.5 vs. 60.0 years, p < .001). The majority of patients in both groups were diagnosed at stage III–IV. A higher proportion of patients treated by PC‐W had residual disease (R1 or R2) than those treated by PC‐3W (41.3% vs. 31.2%, p = .024). Median CA‐125 values at diagnosis were similar in both groups although widely distributed (849 U/mL in the PC‐3W group and 796 U/mL in the PC‐W group, p = .452). BRCA carrier status was tested in 475 patients (67.2%). Of these 475 patients, 317 (66.7%) were noncarriers, 120 (25.3%) were BRCA1 carriers, and 38 (8.0%) were BRCA2 carriers. Patients who were not tested for BRCA status (244/707, 34.5%) were considered BRCA negative in the analyses. A higher rate of patients in the PC‐W group were BRCA1 carriers compared with their rate in the PC‐3W group (22.3% vs. 12.8%, p = .005). The rate of patients who received neoadjuvant chemotherapy was higher in the PC‐W group compared with the PC‐3W with a trend for statistical significance (53.5% vs. 46.9%, p = .083). A similar percentage of patients in the PC‐W group received more than six treatment courses compared with the PC‐3W group (22.7% vs. 20.3%).

The disease recurred in 197 patients (64.6%) in the PC‐W group and in 298 patients (74.1%) in the PC‐3W group (unadjusted HR, .66; 95% CI, 0.55–0.80; p < .001, log‐rank test). After adjustment for age, stage, debulking, BRCA, and histological type, the HR was 0.53 (95% CI, 0.44–0.64; p < .001) in favor of the PC‐W group. Kaplan‐Meier survival analysis showed that median PFS was 21.4 months (95% CI, 18.8–24.2) for the PC‐W regimen and 13.2 months (95% CI, 11.6–16.6) for the PC‐3W regimen (p < .001, Fig. 1). A higher percentage of patients in the PC‐W compared with the PC‐3W group had platinum‐sensitive disease (88.9% vs. 76.3%, p < .001).

During the follow‐up period, 151 patients (49.5%) died in the PC‐W group and 222 (55.2%) died in the PC‐3W group (unadjusted HR, 0.70; 95% CI, 0.57–0.86; p < .001, log‐rank test). After adjustment for age, stage, debulking, BRCA, and histological type, the HR was 0.54 (95% CI, 0.43–0.67; p < .001) in favor of the PC‐W group. Kaplan‐Meier survival analysis showed that median OS was 75.2 months (95% CI, 68.4–86.4) for the PC‐W regimen compared with 54.0 months (95% CI, 48.4–65.7) for the for the PC‐3W regimen (p = .001, Fig. 2).

We also analyzed the population of patients who were ≥75 years at diagnosis (n = 84). Baseline demographic characteristics were similar in both treatment groups (Table 1). In this group of older patients, the disease recurred in 32 patients (70%) in the PC‐W group and in 31 patients (82.0%) in the PC‐3W group (unadjusted HR, 0.63; 95% CI, 0.38–1.04; p = .068, log‐rank test). After adjustment for age, stage, debulking, BRCA, and histological type, the HR was 0.41 (95% CI, 0.22–0.76; p = .005) in favor of the PC‐W group. Kaplan‐Meier analysis showed a median PFS of 14.8 months (95% CI, 9.3–18.1) for the PC‐W regimen and 6.1 months (95% CI, 4.5–9.3) for the PC‐3W regimen (p = .07, Fig. 3).

During the follow‐up period, 27 patients ≥75 years (58.7%) died in the PC‐W group and 30 (79.0%) ‐ in the PC‐3W group (unadjusted HR, 0.54; 95% CI, 0.31–0.93; p = .0236, log‐rank test). After adjustment for age, stage, debulking, BRCA, and histological type, the HR was 0.28 (95% CI, 0.14–0.58; p < .001) in favor of the PC‐W group. Median OS by Kaplan‐Meier survival analysis was 50.8 months (95% CI, 34.0–68.1) for the PC‐W regimen compared with 25.0 months (95% CI, 17.2–32.3) for the PC‐3W regimen (p = .024, Figure 4).

Toxicities

A significantly lower proportion of patients treated by the PC‐W regimen had grade ≥ 2 alopecia compared with those treated with the PC‐3W regimen (39.3%% vs. 88.8%, p < .001, Table 2). The subjective reports on grade 3–4 weakness were lower in the PC‐W regimen compared with the PC‐3W (3.6% vs. 6.2%, p = .15). The rate of grade ≥ 2 neutropenia was significantly lower in the PC‐W group compared with the PC‐3W group (44.9% vs. 53.0%, p = .007), but a similar percentage of patients received GCSF support in both groups (33.7% and 35.8% in the PC‐W and the PC‐3W groups, respectively, p = .570). The rate of anemia grade 3–4 was higher in the PC‐W group compared with the PC‐3W group (12.5% vs 4.5%, p = .001). The rate of thrombocytopenia grade 3–4 was higher for PC‐W than PC‐3W (10.2% vs. 4.5%) but not statistically significantly different. The rate of grade 3–4 neuropathy was significantly lower in the PC‐W group compared with the PC‐3W group (1.6% vs. 5.7%, p = .042). The rate of long‐term peripheral sensory neuropathy after >2 years of follow‐up was also significantly lower in the PC‐W group. None of the patients in the PC‐W group reported grade ≥ 3 nausea compared with 1.2% of patients in the PC‐3W group. In contrast, hypersensitivity reaction was diagnosed in 15 patients (9.7%) in the PC‐W group and in 22 patients (9.4%) in the PC‐3W (p = .937).

Concurrent Treatment

Bevacizumab was approved for OC treatment in Israel from 2014 onward for stage IV at diagnosis or for stage III with residual disease after surgery. In our cohort, only 23.9% and 16.9% of patients in the PC‐W and PC‐3W groups, respectively, were treated with bevacizumab. The 3‐year survival rate was higher in women who were treated concurrently with PC and bevacizumab compared with those who were not treated with bevacizumab (83.0% vs. 76.2%). In woman who were treated concurrently with bevacizumab and PC‐W, the 3‐year survival rate was higher compared with women who were treated concurrently with bevacizumab and PC‐3W (86.3% vs. 79.4%).

Discussion

Safe administration of paclitaxel and carboplatin in patients presenting with advanced ovarian cancer may be challenging because of several factors, including dose, schedule, the need for growth factors, and other supportive care measures, which are important issues in individual patient treatment. The thorough analysis of our institutional experience with the standard paclitaxel and carboplatin double and the weekly regimen adopted from our phase II study because of advantages in OS and tolerance suggests improved outcome and better toxicity profile for a modified PC‐W over the standard PC‐3W regimen.

Table 3 presents a summary of studies that evaluated different schedules of carboplatin‐paclitaxel regimens. The first prospective study depicting the dose‐dense method, in which paclitaxel chemotherapy was administrated at a weekly dose of 80 mg/m² for 6 cycles of 21 days each (with carboplatin AUC6 at day 1), was published in the JGOG 3016 study in 2009 [15], with long‐term follow‐up results published in 2013 [14]. More than 600 women received six cycles of conventional chemotherapy (paclitaxel 180 mg/m² plus carboplatin for an AUC of 6 mg/mL per minute given on day 1 of a 21‐day cycle) or a dose‐dense paclitaxel (80 mg/m² given on days 1, 8, and 15) plus carboplatin given on day 21. Median PFS was longer in the dose‐dense treatment group (28.0 months vs. 17.2 months), as was the 5‐year survival (58.6% vs. 51.0%) [14]. Although withdrawal from the trial was higher in the dose‐dense protocol compared with the conventional protocol, because of toxicity (53% vs 37%), the only significant adverse effect greater in this cohort was grade 3 and 4 anemia. Yet, these results were not replicated in further studies. The American GOG‐0262 study, which included 692 patients [13], the European MITO‐7 study, which included 810 patients who received weekly paclitaxel but not in a dose‐dense method [16], and the ICON‐8 study, which included 1,566 patients [18], did not show any significant prolongation, nor shortening, of the OS or the PFS. The difference between these studies was attributed mainly to pharmacogenomic variations among the research populations, as well as to the changes in paclitaxel doses (Table 3) [19–21]. Notably, in the GOG‐0262 study, among patients who opted not to receive bevacizumab, weekly paclitaxel was associated with a median PFS that was 3.9 months longer than that observed with paclitaxel administered every 3 weeks [13].

Our weekly carboplatin and paclitaxel regimen differs in schedule from others and in particular that of JGOG 3016 [15] and GOG‐262 [13]: we administered carboplatin on days 1, 8, and 15 of a 28‐day cycle at a dose of AUC2, whereas the above “dose‐dense” regimens were given weekly every 21 days. Thus, our protocol allows a “week off” from chemotherapy. In a meta‐analysis [22] that reviewed phase III clinical trials [13, 14, 16], which compared the 3‐weekly and the weekly (dose‐dense) paclitaxel and carboplatin treatment regimens for treatment of ovarian cancer, the authors reported that the dose‐dense regimen was associated with significant improvement of PFS compared with standard schedule, with HR of 0.73 (95% CI 0.61–0.88, p = .001). There was no difference in OS between treatment regimens (HR 0.95, 95% CI, 0.77–1.16, p = .06), as well as in term of severe acute toxicity. These findings corroborate our own experience with the dose‐dense treatment regimen. Some studies also showed an improvement in quality of life and patient satisfaction [16, 17].

As for the toxicity profile, the PC‐3W regimen is known to be toxic, causing alopecia, neurotoxicity, myelosuppression, and fatigue [3, 4, 23]. In the present analysis, the weekly paclitaxel regimen is at least as tolerable as the 3‐weekly one [13, 16]. Although the rates of hematological toxicities were higher in the PC‐W group, this observation might be attributed to the fact that PC‐W–treated patients were followed more closely and were evaluated weekly for toxicity. Therefore, there is a possible underestimation of bone marrow toxicity in the PC‐3W group because additional or midcycle bone marrow toxicity in this group may not have been reported. Of note is the observation that long‐term paclitaxel‐induced peripheral neuropathy was significantly lower in patients treated with the PC‐W regimen. Chemotherapy‐induced peripheral neuropathy is one of the most life‐affecting toxicities of chemotherapy associated with the use of taxanes [24], and it is associated with reduced quality of life in patients with cancer [25, 26]; therefore, reduced long‐term neuropathy in patients treated with PC‐W suggests improved quality of life in these patients.

Although analysis of patient demographics within each one of the treatment regimens (Table 1) suggests that older and more infirm patients received the PC‐W, OS and PFS among patients who were 75 or older at diagnosis also indicate a benefit for treatment with the PC‐W regimen. Better treatment response in PFS and OS of the PC‐W group may be attributed to the 4‐week delay in course completion of the PC‐W regimen (24 weeks instead of 18 weeks for PC‐3W), to the fourth week off‐treatment in each treatment cycle, which allowed the patients to recover, and to the increased completion rate (of 6 cycles) for frail subsets of the population. Thus, our carboplatin weekly regimen with weekly paclitaxel 1, 8, and 15 every 28 days may offer advantages in terms of tolerance in older patients while retaining or improving efficacy, perhaps by allowing more carboplatin to be given.

Our results showed that 3‐year survival rates were somewhat higher for women who were treated concurrently with PC and bevacizumab compared with those who were not, with higher survival rates for the PC‐W regimen at 3 years. The phase‐3 GOG‐0218 study did not find survival differences among patients who received concurrent bevacizumab and PC‐3W compared with PC‐3W alone. A survival benefit was observed in women with BRCA1/2 mutations and in those with non‐BRCA1/2 homologous recombination repair genes [27]. The low percentage of patients in our cohort treated concurrently with carboplatin with paclitaxel and PARP inhibitors prevent us from reaching any conclusion regarding any survival benefit.

Conclusion

Our data suggest that the weekly paclitaxel carboplatin regimen is as active and better tolerated than the standard PC‐3W regimen in a cohort of unselected patients with advanced EOC, TC, and PPC. It is possible that the weekly holiday on day 21 in the PC‐W regimen may ensure better completion rates (that may result in treatment delays for toxicity in PC‐3W). As prospective trials comparing the two regimens are lacking, the results of this analysis highlight the modified dose dense weekly regimen as a valid treatment option, especially for elderly patients and those with significant comorbidities.

Author Contributions

Conception/design: Tamar Safra

Provision of study material or patients: Tamar Safra, Barliz Waissengrin, Talya Levy, Ellie Leidner, Rotem Merose, Diana Matceyevsky, Dan Grisaru, Ido Laskov, Nadav Mishaan, Rotem Shayzaf

Collection and/or assembly of data: Tamar Safra, Barliz Waissengrin, Talya Levy, Ellie Leidner, Rotem Merose, Diana Matceyevsky, Dan Grisaru, Ido Laskov, Nadav Mishaan, Rotem Shayzaf

Data analysis and interpretation: Tamar Safra, Barliz Waissengrin, Ellie Leidner

Manuscript writing: Tamar Safra, Barliz Waissengrin, Ido Wolf

Final approval of manuscript: Tamar Safra, Barliz Waissengrin, Talya Levy, Ellie Leidner, Rotem Merose, Diana Matceyevsky, Dan Grisaru, Ido Laskov, Nadav Mishaan, Rotem Shayzaf, Ido Wolf

Disclosures

The authors indicated no financial relationships.

References

Author notes